Audio fader dB: the rule of thumb for perceived loudness, not a linear percent
A new mixing engineer assumed pulling a fader down '10 dB' meant a 10% volume cut, and was surprised how much quieter the track actually sounded — the dB scale and perceived loudness don't map onto a simple percentage at
We write the way we argue on the bench: exact factor, one sharp trap, one worked example you can re-check.
A new mixing engineer assumed pulling a fader down '10 dB' meant a 10% volume cut, and was surprised how much quieter the track actually sounded — the dB scale and perceived loudness don't map onto a simple percentage at all.
The factor below is the one that ends the argument — write it once, then stop improvising mid-job.
The factor that settles arguments
This is the relationship our converters use. Screenshot it if you have to — just stop re-deriving it from a half-remembered blog post.
Rule of thumb: roughly every -10 dB reduces perceived loudness by about half (a psychoacoustic approximation)
Write that line once at the top of your notes. Every later arithmetic step should point back to it instead of inventing a friendlier number mid-stream.
The mistake that keeps showing up
Reading a dB reduction as if it were a direct percentage cut in volume — dB is logarithmic, and perceived loudness roughly halves for about every 10 dB reduction, which is a much bigger effective cut than '10%.'
I still write the unit next to every intermediate value. It looks pedantic until it catches the one swap that would have shipped wrong.
- Write down the known value in dB reduction.
- Confirm the target unit is really approx. perceived loudness (% of original), not a similar-looking one.
- Apply the factor: Rule of thumb: roughly every -10 dB reduces perceived loudness by about half (a psychoacoustic approximation).
- Round only at the end, to the precision your tool or spec supports.
Worked example
You're pulling a channel fader down by 10 dB in a mix and want a sense of how much quieter that will actually sound.
A -10 dB fader move: perceived loudness drops to roughly 50% of the original, not 90% as a naive 'percent' reading of the dB number might suggest.
That's roughly half as loud at -10 dB — this is a psychoacoustic rule of thumb, not an exact physical law, and doesn't apply the same way to every listener or frequency.
If you want a second check, invert the conversion and see whether you recover the original quantity within normal rounding.
Quick reference table
| dB reduction | Calculation | approx. perceived loudness (% of original) |
|---|
Use the table for speed once you trust the factor. Do not use it as a substitute for understanding which definition of the unit you are on.
Why the exact number matters
Mixing decisions rely on understanding that small dB moves have outsized perceived effects — misjudging this leads to overcorrecting or undercorrecting a track's level in a mix.
Precision is not about showing off decimals. It is about making sure the shopping list, the machine setting, and the acceptance check are all describing the same physical quantity.
Pocket recap
Treat roughly every 10 dB of reduction as cutting perceived loudness in about half, not as a percentage of the dB number itself — it's a rule of thumb, not an exact conversion.
Keep the converter link nearby for the next time the same debate shows up in a group chat. Re-typing the factor from memory is how the wrong cousin of a unit sneaks back in.
Try the converters mentioned in this article
FAQ
What is the exact dB reduction to approx. perceived loudness (% of original) factor here?+
Rule of thumb: roughly every -10 dB reduces perceived loudness by about half (a psychoacoustic approximation)
Should I round partway through the calculation?+
No — carry full precision through the multiplication and round only the final answer, to the precision your tool, gauge, or spec actually supports.
What's the single biggest mistake people make with this conversion?+
Reading a dB reduction as if it were a direct percentage cut in volume — dB is logarithmic, and perceived loudness roughly halves for about every 10 dB reduction, which is a much bigger effective cut than '10%.'
This article was written by The Turbo Unit Converter engineering desk (Applied units & field metrology) and last reviewed on 2026-07-25 against NIST SP 811 and the BIPM SI Brochure. Read our full editorial policy.